Just add a standard curve in your gpcr and read the value out directly (quantitaive real time pcr).

It would help to know what your "problem" is. If you have a sample where you know that the target and control sequences are equimolar, then using that sample to construct a standard curve as Godfrey suggested is a good idea. If you do not have such a control sample, then you will need to make a control where you know that the two sequences are equimolar (for example, an equimolar mixture of purified plasmids or PCR products). I'm not sure how you determined PCR efficiency if you didn't have a control mixture.....

To determine efficiency I started from a sample of genomic DNA as template and run a first PCR with each primer set separately. Then, I serially diluted each PCR product and performed a qPCR reaction with the respective primer set, determining Ct value for each dilution and then plotting Ct vs Log[DNA].

Why is important to have a sample with equimolar amounts of target sequences?

The abundance of each DNA in my samples are very distinct. I'm using this approach to determine the excision frequency of a genomic island (one primer set amplify the scar that leaves the island after excision, and the other set a gene that is in one copy per chromosome), so the deal is to calculate the number of chromosomes that have the scar divided by the total of chormosomes, in a sample of genomic DNA. The expected frequency is about 10^-5. So, to get such an equimolar control, I must construct it from mixing PCR amplicons.

Thanks for your help!

On second thought, the best control would be to get an isolate where the genomic island has already excised--i.e. that contains the scar and rpoD . If you have such an isolate, then you can use DNA from it and be guaranteed that the scar and rpoD are "equimolar". I don't know where the island and rpoD are relative to each other on the chromosome, but depending upon where they are, their absolute copy numbers could differ by 2 or 3 powers of 2 within a cell, assuming that your bacteria are growing in rich medium. So, in principle, this could make an order-of-magnitude difference in your estimate of excision rates, and this could be affected by mutants (I assume that you are checking different mutants for excision rates....). So even if you have to hunt through 10^5 colonies, it might be worth trying to get a pre-excised isolate.

Well, I agree with you in that the "cured" isolate would be the best control. Sadly we don't have this strain yet.

Please, could you explain me why the location of the island and rpoD relative to each other on the chromosome can produce such differences in the absolute copy number within a cell?. We are interested in comparing the excision rates in different conditions (phase of growth, overexpression of related mobility genes, etc.).

Assuming that this is E. coli, there can be as many as 3 replication forks per chromosome, so the ratio of the origin to terminus can be as high as 8.

In this case relative quantitative approach is suitable. you can run PCR in triplicate for each sample to calculate exact copy number on genomic sample. If you do standard curve without samples in PCR set, you would have different efficiency in PCR set. you can obtain PCR efficiency without running standard curve by using REST2009 software. another point is that standard curve is most useful in absolute quantification.

In real-time PCR, triplicated amplification Curve of each gene (target and reference) should be overlapped well with each other. Analysis of quantitative real-time PCR data will obtain by comparative threshold cycle method (ΔΔCt) and measuring of Target/reference genes ratio using (the Relative Quantitative approach) the 2-ΔΔCt formula (Target/reference ratio = 2-ΔΔCt) (95% Confidence Interval for gene dosage) in normal subjects and case, respectively.